Improved understanding of the potential growth of urban areas at the national and global levels is needed for sustainable urban development. Current panel data analysis and local scale modeling are limited in projecting global urban area growth with large spatial heterogeneities. In this study, we developed country-specific urban area growth models using the time series data set of global urban extents (1992–2013) and projected the future growth of urban areas under five Shared Socioeconomic Pathways (SSPs). Our results indicate the global urban area would increase roughly 40–67% under five SSPs until 2050 relative to the base year of 2013, and this trend would continue to a growth ratio of more than 200% by 2100. The growth of urban areas under relatively unsustainable development pathways (e.g., regional rivalry SSP3 and inequality SSP4) is smaller compared to other SSPs. Although developing countries would remain as leading contributors to the increase of global urban areas in the future, they may exhibit different temporal patterns, that is, plateaued or monotonically increasing trends. This variation is primarily attributed to the compounding effect of the growth in population and gross domestic product. Our urban area data set presents a first country-level urban area projection under the five SSPs, spanning from 2013 to 2100. This data set has a great potential to support various global change studies, for example, urban sprawl simulation, integrated assessment modeling for sustainable development goals, and investigation of the impact of urbanization on atmospheric emissions, air quality, and human health.
- We projected urban area growth until 2100 in each country by developing country-specific urban growth models
- Global urban area would increase by 40–67% under five SSPs until 2050 with large variations among countries
- The generated data set is the first country-level urban extents under five SSPs with potential use in studies of global environmental changes
Plain Language Summary
Improved understanding of the potential growth of urban areas at the national and global levels is highly needed for sustainable urban development. In this study, we developed country-specific urban area growth models using the time series data set of global urban extents (1992–2013) and projected the future growth of urban areas under five Shared Socioeconomic Pathways (SSPs). We found the global urban area would increase by roughly 40–67% under five SSPs until 2050 relative to the base year of 2013, and this trend would continue to a growth ratio of more than 200% by 2100. Although developing countries would remain as leading contributors to the increase of global urban areas in the future, they may exhibit different temporal patterns, that is, plateaued or monotonically increasing trends. Our urban area data set presents a first country-level urban area projection under the five SSPs, spanning from 2013 to 2100. This data set has a great potential to support various global change studies, for example, urban sprawl simulation, integrated assessment modeling for sustainable development goals, and investigation of the impact of urbanization on atmospheric emissions, air quality, and human health.
Our planet has been experiencing rapid urbanization over past decades, which changes the urban environment and affects human health, and further challenges the realization of sustainable development goals (SDGs). According to the latest World Urbanization Prospects (United Nations, 2018), more than 66% of the global population is projected to reside in urban areas by 2050. Moreover, the urbanization rate is higher in developing regions, such as countries in Africa and Asia, where many people are migrating from rural to urban, resulting in a notable expansion of urban extent and other concerns about energy and urban environment (Alberti et al., 2017; Güneralp et al., 2017; Li et al., 2016). As one of the key indicators of urbanization, the rapid global urban sprawl poses a wide range of challenges to the sustainable development at the local, national, and global scales. For example, it adversely affects forest and agricultural ecosystems and services they provide, such as food, fiber, and energy for humans (DeFries et al., 1999; Foley et al., 2005). Rapid urbanization and urban sprawl can also intensify urban heat island effect and increase building energy consumption (Imhoff et al., 2010; Zhang et al., 2013; Zhou, Clarke, et al., 2014). Also, it can negatively affect the urban environment by raising concerns about air pollution and public health (Gong et al., 2012; Li et al., 2019) and can result in drastic land use conversion and then significantly influence ecosystem services (Wang et al., 2019). Therefore, understanding the ongoing urbanization and quantifying its pace and magnitude going forward is crucial to sustainable development (Lu et al., 2015).
It is challenging to expand local-scale urban area growth modeling to other regions or use them for the projection of global urban area growth because of the differences such as methods and socioeconomic drivers used in these studies (Santé et al., 2010). For example, He et al. (2016) predicted the increase of urban areas in Beijing (China) through 2030 using population data. The urban area growth model was built on a simple linear regression model, in which the projected population primarily determined the growth of urban areas. However, the relationship between urban areas and socioeconomic indicators (e.g., population and gross domestic product - GDP) is treated differently in the literature (Santé et al., 2010). Approaches used for urban area projection range from directly coupling with the estimated population growth rate (Li & Yeh, 2000) or employing the urban area planned by the government (Chen et al., 2002) to adopting ancillary socioeconomic indicators such as per capita GDP or income (Li et al., 2014; Wu et al., 2010) or using more sophisticated system dynamics models (Han et al., 2009). Overall, the derived city (or local) scale relationships are not applicable to cities with different urbanization levels (Irwin & Geoghegan, 2001), and the required data sets in these city-scale models are not widely available at the regional or global scale.
Most studies on global urban area growth do not consider historical pathways of urban areas at the country (or regional) scales. Seto et al. (2012) estimated the increase of global urban areas using a Monte Carlo approach, which was implemented based on a global urban extent map (ca. 2000) and the estimated regional-scale population densities. However, historical pathways of urban area growth and spatially different urbanization levels were not considered. Klein Goldewijk et al. (2010) used the local urban density (i.e., per capita urban areas) to project the growth of urban areas at the global scale from 10,000 BCE to 2005. However, the resolution of resulting urban map is relatively coarse (around 8 km), and the model assumes an asymmetric bell-shaped curve to represent the temporal change of urban density only using a limited number of surveys from cities published in literature, which introduces high uncertainties in overall analysis (Dong et al., 2018; Hurtt et al., 2006; Li et al., 2016, 2017). Limited knowledge of the dynamics of urban areas over past decades has been the primary challenge in understanding the historical pathway of urban sprawl and projecting its future growth at the national and global scales (Li & Gong, 2016; Li et al., 2018).
This study aims to provide a global projection of urban area growth in the future for all countries. Unlike previous studies that were conducted at the local scale or ignored historical pathways of urban sprawl in each country, we developed country-specific urban area growth models using the time series (22 years) data of urban areas from satellite observations and projected global urban area growth in future, under different socioeconomic pathways. The remaining of this paper is organized as follows. Sections 2 and 3 provide descriptions of the data set and methodology, respectively. Section 4 analyzes projected growth of urban areas for individual countries and the global. Section 5 presents conclusions from this study and some directions for future studies.
2.1 Historical Annual Urban Extents
One of the major data sets used in this study is the annual dynamics of global urban extent from 1992 to 2013, derived from the nighttime light data (Zhou et al., 2018). This data set differs from other existing nighttime light-based studies by mapping the urban extent at both regional and global scales (Small et al., 2005; Zhou, Smith, et al., 2014; Zhou et al., 2015). This data set is also unique because of its spatial and temporal consistency and a good agreement with other existing independent urban extent products (Li et al., 2015; Zhou et al., 2018). The derived urban extent data set has a 1-km spatial resolution and an annual temporal resolution, spanning from 1992 to 2013. The mean accuracies of this data set in the United States, China, and Europe are 96%, 86%, and 95%, respectively, at the urban cluster level according to the fine-resolution land cover data. This new data set depicts cities that have different urbanization rates in different regions. For example, cities in developing areas such as in New Delhi (India) and Beijing (China) show a more notable expansion of urban extent compared to cities in developed regions, such as Atlanta (USA; Figure 1).
There is unprecedented and rapid urbanization worldwide over the past two decades, although the rate and extent of urbanization vary considerably across countries and years (Figure 2). Compared to the base year of 1992, most countries experienced more than a doubling of urban area growth by 2013. The majority of rapidly expanding urban areas are located in Asia, Africa, and South America, which appear to be the dominant driver of global urbanization in the future. At the same time, countries in these rapidly developing regions are likely to face more challenges for continuing the desired urbanization trends (Solecki et al., 2013), that is, achieving a balance between expanding the access to resources (such as fresh water, energy, and affordable and nutritious food) and minimizing adverse impacts on ecosystems and surrounding environment.
2.2 Socioeconomic Data
We used the historical records of socioeconomic data (i.e., population and GDP) to develop country-specific urban area growth models. Time series data of annual population and nominal GDP (1992–2013) were obtained from the World Bank database (http://databank.worldbank.org/). To make the GDP data comparable across countries and over time due to changes in living standards, we adjusted the nominal GDP to real GDP using the official rates of purchasing power parity and deflation. The resulting historical socioeconomic data (i.e., population and GDP; 1992–2013) were then used to develop urban area growth model for each of 171 countries used in this study, except for countries without records in the World Bank database.
We used future socioeconomic indicators derived from the Shared Socioeconomic Pathway (SSP) database to project the growth of urban area during 2013–2100. The SSP database is from the International Institute for Applied System Analysis (https://tntcat.iiasa.ac.at/SspDb/), which quantifies possible pathways of socioeconomic development in the future (Riahi et al., 2017). There are five different socioeconomic narratives in the SSP database, which span a range of challenges to mitigation of and adaptation to climate change (Table 1). These narratives were translated into different pathways of socioeconomic development as they present various quantitative scenarios for national (regional) population and GDP change in the future (Figure 3). For example, SSP1 (sustainability) has the slowest pace of population growth and relatively rapid GDP growth, which is opposite to the SSP3 (region rivalry) scenario (Dellink et al., 2017; Kc & Lutz, 2017).
|SSP2||Middle of the road||Medium||Medium|
|SSP5||Fossil fueled development||High||Low|
- Note. SSP = Shared Socioeconomic Pathway.
We developed country-specific urban area growth models using the time series data of global urban extents (1992–2013) and then projected the future growth of urban area under five different SSPs (Figure 4). First, country-specific urban area growth models were developed based on the time series of per capita cumulative GDP and per capita urban areas, using a proposed sigmoid growth function. Second, future urban area growths under different socioeconomic pathways were projected using the population and GDP data from the SSP database, throughout the century (2013–2100). These two procedures are described in details in the following sections.
3.1 Country-Specific Urban Area Growth Models
The proposed sigmoid growth model can well capture different phases of urban development (i.e., initial, middle, and mature) in different countries (Figure 5). Trajectories of urban area growth at different urbanization phases were obtained from the annual time series data (Figure 5a). For countries at the initial phase of urbanization (e.g., countries in North Africa or middle Asia), the growth of per capita urban areas remains slow relative to the growth of per capita cumulative GDP. For countries experiencing a midphase of urbanization stage (e.g., emerging economies such as China or India), the per capita urban areas grow faster, possibly due to the industrialization that involves the development of urban lands. For countries at a mature phase of urbanization (e.g., countries in North America or Europe), the growth of per capita urban areas levels off at a high and stable level. The proposed sigmoid growth model can capture the relationship between the per capita urban areas and the per capita cumulative GDP at different phases of urbanization, using the long time series of urban areas. Compared to the multiregression model (Dong et al., 2018; He et al., 2013), which was generally performed in specific regions or countries, the sigmoid growth model only requires a constant set of parameters at different phases of urbanization. Although such stage-wise response of per capita urban areas to per capita cumulative GDP is observed in most countries, the fitted sigmoid curves vary by country as presented in Figure 5b. For example, curves of rapidly developing countries (e.g., China and India) are notably different from that of developed countries (e.g., United States and Canada). More cases can be found in Figure S1. These results are consistent with previous literature that report urbanization phases in different countries have different growth patterns, that is, a rapid expansion of urban areas in developing countries (i.e., midphase of urbanization) and a relatively stable (or steady) growth in developed countries (i.e., mature phase of urbanization; Eom et al., 2012; Liu & Phinn, 2003; Yu et al., 2014).
3.2 Urban Area Projection
4 Results and Discussion
4.1 Performance of Country-Specific Urban Area Growth Models
The proposed urban area growth model performs well for most countries in capturing historical urban area growth (Figure 6). The overall performance, as measured by the goodness-of-fit between the modeled and the observed data, is good with R-square values greater than 0.8 in most countries (Figure 6). This suggests that the proposed sigmoid growth model properly captures the historical pattern of urban sprawl. For several countries, however, the model performs relatively poor, and these countries are mainly located in middle Africa in which the urbanization level is relatively low. The lower quality of reported population and GDP data for these countries is the likely reason for this poor performance of the proposed urban area growth model (United Nations, 2018).
4.2 Global Urban Area Growth
The projected urban area growth is different among the five SSPs used in this study, resulting in a notable difference of urban areas across SSPs in the second half of the century. All SSPs indicate a continued increase in urban areas under a range of development pathways (Figure 7), although the growth rate varies across scenarios. Projections under SSP3 and SSP4 are noticeably lower compared to SSP1, SSP2, and SSP5. This difference can be explained by different development narratives that influence population and GDP growth trajectories. For example, SSP3 represents the regional rivalry scenario, in which the income growth is not as fast as the population growth; SSP4 is the inequality scenario with highly unequal economic opportunities and political power across and within regions, resulting in the relatively slow growth of population and GDP (O'Neill et al., 2017). Thus, per capita urban areas under these two scenarios remain relatively low and so is the increase of total urban areas. Although SSP3 and SSP4 show similar GDP increments in the future, the population growth in SSP3 is notably higher than SSP4 (Figure 3), resulting in a lower per capita cumulative GDP in SSP3. Thus, the per capita urban area is also lower under SSP3 according to the sigmoid growth model, while, the lower per capita urban area and higher population under SSP3 together result in a similar total urban area as SSP4. Also, it should be noted that the global urban area growths under five SSPs were derived from aggregation of all countries, which have their own pathways. By contrast, under the other three scenarios (i.e., SSP1, SSP2, and SSP5) that exhibit a consistent rise in GDP and limited growth in population by the end of this century (Figure 2), a considerable growth of urban areas is observed, which is determined by the increment of both the per capita cumulative GDP and the total population. A medium level of growth in urban areas is found under the SSP2, consistent with its middle-of-the-road development projection (Figure 7). Also, the five SSPs all have a modest urban area growth until the midcentury before they take increasingly different growth pathways (Table 2). For example, urban growth rates for the five SSPs remain relatively low and similar until 2020 (around 6.8%), but they diverge as much as 27% by the midcentury (between SSP5 and SSP4), and almost 100% by the century end (between SSPs 1, 2, or 5 and SSPs 3 or 4).
- Note. SSP = Shared Socioeconomic Pathway.
There is a distinctive spatial difference in urban area growth among the five SSPs. Regardless of the SSPs, the growth rate in developing areas, such as Africa and Asia, stays higher than that of developed regions, such as North America or Europe (Figures 8 and 9). This pattern is consistent with the trend over the past two decades (Figure 5). Besides, the expansion of urban areas in Asia and Africa is more pronounced than that of other continents. In particular, several countries, such as India or Nigeria, in which populations are projected to increase rapidly during the century (United Nations, 2018; Jiang & O'Neill, 2017), achieve almost doubled growth rate of the other regions in all SSPs, particularly in the second half of the century. Nevertheless, there is still a large gap to close before they reach the level of per capita urban areas of developed countries. By contrast, the developed regions, such as North America and Europe, exhibit a relatively modest increase in urban areas, with its growth rate being less than half of developing regions. Additionally, the developed regions exhibit rather significant growth in urban areas only under SSP5 due to its rapid economic development (Figures 8 and 9; Riahi et al., 2017).
4.3 Urban Area Growth in Major Countries
The growth of urban area shows noticeably different patterns across five SSPs in major countries, even though these countries currently are in similar urbanization phases. To better understand the influence of socioeconomic indicators on the growth of urban areas, we selected three major countries (i.e., the United States, China, and India) at different urbanization levels for further investigation (Figure 10). The United States is a highly urbanized country, and China and India are two rapidly developing countries each hosting more than 1.3 billion people. The total urban area of these three countries accounts for about 30% of the global urban area. Unlike the United States or China, which show a relatively slow growth or even a decrease in their population expected in the near future (ca. 2030), the population in India is likely to increase until 2050, totaling 1.7 billion people according to SSP2, the middle-of-the-road development pathway (O'Neill et al., 2017). Taking SSP2 as an example, although per capita cumulative GDP and per capita urban areas of these three countries would continue to rise, the increment of per capita urban areas in the United States is plateaued earlier compared to other two developing countries: China and India. Meanwhile, the growth of China's per capita urban areas would level off after 2050, which is distinctively different from India that would continue to increase (Figure 10). Considering the population dynamics in China and the United States, to a lesser degree, they would exhibit a relatively modest urban area growth particularly in the second half of this century. Our results confirm that the United States shows a relatively slow increase in urban areas (a 48% growth compared to 2013) throughout the century, which is considerably lower than that of developing countries, such as China (175%) and India (800%; Figure 9). In SSP5, a fossil fuel-based development pathway, highly developed countries like the United States would likely to experience a continued increase in urban areas due to the rapid economic development combined with intensive consumption of resources (O'Neill et al., 2017). These different pathways of urban area growth under the five SSPs can provide a comprehensive and quantitative data set for spatially explicit modeling of urban sprawl and evaluating progress toward achieving the SDGs (Lu et al., 2015).
In this study, we developed country-specific urban area growth models using the time series of global urban extent and historical socioeconomic indicators (i.e., population and GDP) over the past two decades (1992–2013). The developed urban area growth models show a reasonably good performance for most countries worldwide. Our model is based on a long time series urban area data and can well capture different phases of urban development, which is notably different from the existing urban area analysis conducted using limited panel data observations. These models were used to project urban areas for different regions and countries for the rest of this century under five sets of population and GDP scenarios derived from SSPs. Our projections provide the first of its kind country-level data set of urban areas with due consideration to both historical urban growth pathways and future SSP scenarios, spanning from 1992 to 2100 (i.e., both retrospective and prospective) in a globally consistent manner.
Our projections reveal different pathways of urban area growth under the five SSPs across different countries, which is essential for investigating spatially explicit impacts of future urban sprawl on society and ecology and for achieving SDGs. There would be approximately 40–67% increase in urban areas globally under five SSPs by the midcentury, relative to the base year of 2013. The pace of urban area growth would be amplified from 2050 to 2100, possibly presenting doubling or tripling of urban areas depending on development pathways. Urban area growth under development pathways such as the regional rivalry scenario (SSP3) and the inequality scenario (SSP4) is constrained compared to other SSPs, such as the fossil fuel-based development scenario (SSP5) and the sustainable development scenario (SSP1). Also, the rapid growth of socioeconomic development in countries like China and India would continue to be the predominant driver of global urban area growth. Moreover, for some developing countries, such as China, urban area growth would level off around 2050, which is considerably different from the majority of developing countries such as India.
The projections of urban areas would be useful for global urban sprawl modeling and sustainable development studies. The top-down urban areas projected in this study can be coupled with a bottom-up spatial allocation model such as cellular automata (Li et al., 2014; Li & Gong, 2016), to simulate spatially explicit sprawl of urban extent. Besides, our SSP-variant urban area data set provides multiple urban development pathways that span a wide range of future challenges associated with climate change. Thus, human and ecosystem vulnerability and resilience to climate change and other environmental stressors can be assessed under various realizations of future socioeconomic development. This data set can also be used to explore potential risks (e.g., biodiversity hotspots) caused by urban sprawl (Myers et al., 2000), assess the challenges faced by individual countries in their efforts to achieve the SDGs and associated national targets (Güneralp et al., 2017; Seto et al., 2012), and develop robust responses and management of resources at the national and regional level. Additionally, this data set can be used with Earth system models to explore the anthropogenic impacts on water, energy and carbon cycles, and global environment (Bond-Lamberty et al., 2014; Collins et al., 2015). However, it should be noted that a more comprehensive assessment of global environmental changes would also consider a detailed representation of other quantitative inputs (e.g., natural resources, technologies, and investments) and qualitative narratives (e.g., policies, institutions, lifestyle, and settlement pattern) that are consistent with the major SSP indicators used in this study (i.e., population and GDP; van Vuuren et al., 2017). For example, although the total urban area under SSP1 (Sustainability) shows a twofold increase by 2100 relative to the base year of 2013, the new urban area mostly comes from developing countries in reducing inequality within and across countries for the sustainable development (Figure 10). Thus, future studies on urban area dynamics would benefit greatly in identifying plausible sustainable development pathways, especially for developing regions and nations worldwide.
We acknowledge funding support from the U.S. Department of Energy, Office of Science, as part of research in Multi-Sector Dynamics, Earth and Environmental System Modeling Program, the NASA ROSES LULC Program “NNH11ZDA001N-LCLUC,” and the NASA ROSES INCA Program “NNH14ZDA001N-INCA,” with additional support for Jiyong Eom from the Ministry of Environment of Korea through the Climate Change Correspondence Program (2004001300002). The historical and future population and GDP data are available in the World Bank Database (http://databank.worldbank.org/) and the Shared Socioeconomic Pathways (SSP) Database (https://tntcat.iiasa.ac.at/SspDb), respectively. The projected global urban area growth under five SSPs by 2100 is available at figshare website (https://doi.org/10.6084/m9.figshare.7817624.v1).We would like to thank two anonymous reviewers for their constructive comments and suggestions. The views and opinions expressed in this paper are those of the authors alone.
|eft2527-sup-0001-2019EF001152-SI.docxWord 2007 document , 162.2 KB||Supporting Information S1|
Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.
- 2017). Global urban signatures of phenotypic change in animal and plant populations. Proceedings of the National Academy of Sciences of the United States of America, 114, 8951–8956. https://doi.org/10.1073/pnas.1606034114
- 2014). On linking an Earth system model to the equilibrium carbon representation of an economically optimizing land use model. Geoscientific Model Development, 7, 2545–2555. https://doi.org/10.5194/gmd-7-2545-2014
- 2002). Assessment of the urban development plan of Beijing by using a CA-based urban growth model. Photogrammetric Engineering and Remote Sensing, 68, 1063–1072.
- 2015). The integrated Earth system model version 1: Formulation and functionality. Geoscientific Model Development, 8, 2203–2219. https://doi.org/10.5194/gmd-8-2203-2015
- 1999). Combining satellite data and biogeochemical models to estimate global effects of human-induced land cover change on carbon emissions and primary productivity. Global Biogeochemical Cycles, 13, 803–815.
- 2017). Long-term economic growth projections in the Shared Socioeconomic Pathways. Global Environmental Change, 42, 200–214. https://doi.org/10.1016/j.gloenvcha.2015.06.004
- 2018). Land use projections in China under global socioeconomic and emission scenarios: Utilizing a scenario-based land-use change assessment framework. Global Environmental Change, 50, 164–177. https://doi.org/10.1016/j.gloenvcha.2018.04.001
- 2012). China's building energy use: A long-term perspective based on a detailed assessment, PNNL Report, PNNL-21073.
- 2005). Global consequences of land use. Science, 309, 570–574. https://doi.org/10.1126/science.1111772
- 2012). Urbanisation and health in China. Lancet, 379, 843–852. https://doi.org/10.1016/S0140-6736(11)61878-3
- 2017). Global scenarios of urban density and its impacts on building energy use through 2050. Proceedings of the National Academy of Sciences of the United States of America, 114, 8945–8950. https://doi.org/10.1073/pnas.1606035114
- 2009). Application of an integrated system dynamics and cellular automata model for urban growth assessment: A case study of Shanghai, China. Landscape and Urban Planning, 91, 133–141. https://doi.org/10.1016/j.landurbplan.2008.12.002
- 2016). Assessing the potential impacts of urban expansion on regional carbon storage by linking the LUSD-urban and InVEST models. Environmental Modelling and Software, 75, 44–58. https://doi.org/10.1016/j.envsoft.2015.09.015
- 2013). Modeling the urban landscape dynamics in a megalopolitan cluster area by incorporating a gravitational field model with cellular automata. Landscape and Urban Planning, 113, 78–89. https://doi.org/10.1016/j.landurbplan.2013.01.004
- 2006). The underpinnings of land-use history: Three centuries of global gridded land-use transitions, wood-harvest activity, and resulting secondary lands. Global Change Biology, 12, 1208–1229. https://doi.org/10.1111/j.1365-2486.2006.01150.x
- 2010). Remote sensing of the urban heat island effect across biomes in the continental USA. Remote Sensing of Environment, 114, 504–513. https://doi.org/10.1016/j.rse.2009.10.008
- 2001). Theory, data, methods: Developing spatially explicit economic models of land use change. Agriculture, Ecosystems & Environment, 85, 7–24. https://doi.org/10.1016/S0167-8809(01)00200-00206
- 2017). Global urbanization projections for the Shared Socioeconomic Pathways. Global Environmental Change, 42, 193–199. https://doi.org/10.1016/j.gloenvcha.2015.03.008
- 2017). The human core of the shared socioeconomic pathways: Population scenarios by age, sex and level of education for all countries to 2100. Global Environmental Change, 42, 181–192. https://doi.org/10.1016/j.gloenvcha.2014.06.004
- 2010). Long-term dynamic modeling of global population and built-up area in a spatially explicit way: HYDE 3.1. The Holocene, 20, 565–573. https://doi.org/10.1177/0959683609356587
- 2017). A new global land-use and land-cover change product at a 1-km resolution for 2010 to 2100 based on human–environment interactions. Annals of the American Association of Geographers, 107, 1040–1059. https://doi.org/10.1080/24694452.2017.1303357
- 2016). Urban growth models: Progress and perspective. Scientific Bulletin, 61, 1637–1650. https://doi.org/10.1007/s11434-016-1111-1
- 2015). A 30-year (1984–2013) record of annual urban dynamics of Beijing City derived from Landsat data. Remote Sensing of Environment, 166, 78–90. https://doi.org/10.1016/j.rse.2015.06.007
- 2014). A systematic sensitivity analysis of constrained cellular automata model for urban growth simulation based on different transition rules. International Journal of Geographical Information Science, 28, 1317–1335. https://doi.org/10.1080/13658816.2014.883079
- 2000). Modelling sustainable urban development by the integration of constrained cellular automata and GIS. International Journal of Geographical Information Science, 14, 131–152.
- 2016). A cellular automata downscaling based 1 km global land use datasets (2010–2100). Scientific Bulletin, 61, 1651–1661. https://doi.org/10.1007/s11434-016-1148-1
- 2019). Characterizing the relationship between satellite phenology and pollen season: A case study of birch. Remote Sensing of Environment, 222, 267–274. https://doi.org/10.1016/j.rse.2018.12.036
- 2018). Mapping annual urban dynamics (1985-2015) using time series of Landsat data. Remote Sensing of Environment, 216, 674–683. https://doi.org/10.1016/j.rse.2018.07.030
- 2003). Modelling urban development with cellular automata incorporating fuzzy-set approaches. Computers Environment and Urban, 27, 637–658. https://doi.org/10.1016/S0198-9715(02)00069-8
- 2015). Five priorities for the UN sustainable development goals. Nature, 520, 432–433.
- 2013). Detecting interannual variation in deciduous broadleaf forest phenology using Landsat TM/ETM + data. Remote Sensing of Environment, 132, 176–185. https://doi.org/10.1016/j.rse.2013.01.011
- 2015). Detecting change in urban areas at continental scales with MODIS data. Remote Sensing of Environment, 158, 331–347. https://doi.org/10.1016/j.rse.2014.09.023
- 2000). Biodiversity hotspots for conservation priorities. Nature, 403, 853. https://doi.org/10.1038/35002501, https://www.nature.com/articles/35002501
- 2017). The roads ahead: Narratives for shared socioeconomic pathways describing world futures in the 21st century. Global Environmental Change, 42, 169–180. https://doi.org/10.1016/j.gloenvcha.2015.01.004
- 2017). The Shared Socioeconomic Pathways and their energy, land use, and greenhouse gas emissions implications: An overview. Global Environmental Change, 42, 153–168. https://doi.org/10.1016/j.gloenvcha.2016.05.009
- 2010). Cellular automata models for the simulation of real-world urban processes: A review and analysis. Landscape and Urban Planning, 96, 108–122. https://doi.org/10.1016/j.landurbplan.2010.03.001
- 2012). Global forecasts of urban expansion to 2030 and direct impacts on biodiversity and carbon pools. Proceedings of the National Academy of Sciences of the United States of America, 109, 16,083–16,088. https://doi.org/10.1073/pnas.1211658109
- 2005). Spatial analysis of global urban extent from DMSP-OLS night lights. Remote Sensing of Environment, 96, 277–291. https://doi.org/10.1016/j.rse.2005.02.002
- 2013). It's time for an urbanization science. Environment: Science and Policy for Sustainable Development, 55, 12–17. https://doi.org/10.1080/00139157.2013.748387
United Nations (2018). World urbanization prospects: The 2018 revision, UN.
- 2017). The Shared Socio-economic Pathways: Trajectories for human development and global environmental change. Global Environmental Change, 42, 148–152. https://doi.org/10.1016/j.gloenvcha.2016.10.009
- 2019). A multiscale analysis of urbanization effects on ecosystem services supply in an urban megaregion. Science of the Total Environment.
- 2010). Simulating urban expansion by coupling a stochastic cellular automata model and socioeconomic indicators. Stochastic Environmental Research and Risk Assessment, 24, 235–245.
- 2014). Scenarios of building energy demand for China with a detailed regional representation. Energy, 67, 284–297. https://doi.org/10.1016/j.energy.2013.12.072
- 2013). Energy consumption and the unexplained winter warming over northern Asia and North America. Nature Climate Change, 3, 466–470. https://doi.org/10.1038/nclimate1803
- 2014). Modeling the effect of climate change on U.S. state-level buildings energy demands in an integrated assessment framework. Applied Energy, 113, 1077–1088. https://doi.org/10.1016/j.apenergy.2013.08.034
- 2018). A global record of annual urban dynamics (1992-2013) from nighttime lights. Remote Sensing of Environment, 219, 206–220.
- 2014). A cluster-based method to map urban area from DMSP/OLS nightlights. Remote Sensing of Environment, 147, 173–185. https://doi.org/10.1016/j.rse.2014.03.004
- 2015). A global map of urban extent from nightlights. Environmental Research Letters, 10, 1–11. https://doi.org/10.1088/1748-9326/10/5/054011